Last month, hours after the announcement of QEternity, we said [7] that in validation of the 'Flow' model taking over from the Fed's flawed 'Stock' model, the Fed will have no choice but to continue the long-end $85 billion in monetary flow addition to the market, if not economy (i.e. expand the QE program from $40bn per month to $85bn per month starting in January - in order to maintain the 'flow' post-Operation Twist). Last night, Goldman has officially agreed with us (as has Bloomberg's chief economist Joe Brusuelas). It appears that starting January 2013 Ben is really going to town. But don't expect this to be announced today. It will, as Goldman speculates, be disclosed at the Fed's December FOMC meeting. For now, two weeks ahead of the election, expect more "autopilot" from Bernanke as coming up with any surprises 'now' would be seen as beyond political.
Via Goldman Sachs' Jan Hatzius: Waiting For December
- We expect virtually no changes in Wednesday's FOMC statement, except for small tweaks updating the description of recent economic data and recognizing that QE3 has now begun.
- The December 11-12 meeting will be much more interesting. We expect the QE program to expand from $40bn per month now to $85bn per month from January, in order to keep the pace of asset purchases constant even after the completion of Operation Twist. A shift to outcome-based forward guidance (instead of the current "mid-2015" calendar-based guidance) and/or adoption of an FOMC consensus forecast for the economy are also under discussion, although we are not forecasting either at this point.
Q: What do you expect from the FOMC statement on Wednesday?
A: Very little. The first paragraph describing the economy might see a few small changes, but the only one that is clearly necessary is a reformulation of the sentence on inflation. The sentence currently notes that "[i]nflation has been subdued, although the prices of some key commodities have increased recently." Since September 13, headline inflation has been higher but core inflation remains low and commodity prices have stabilized, so we might see a reformulation along these lines. The committee might also upgrade its assessment of the labor and/or housing market slightly, possibly noting the decline in the unemployment rate or the faster recent pace of improvement in the housing indicators. That said, we do not expect them to alter their assessment that the unemployment remains "elevated." We do not expect any meaningful changes to the remainder of the statement, although the language is likely to be adjusted in order to recognize that QE3 has now begun.
Q: What about the December 11-12 meeting?
A: This will be a much more interesting meeting. The most immediate question is the pace of asset purchases after Operation Twist (or the Maturity Extension Program in Fed parlance) concludes at the end of December. We believe that the committee will be reluctant to do anything that markets would interpret as a slowdown in the pace of monetary easing, such as an indication that the pace of asset purchases will slow in the foreseeable future. Our baseline expectation is a continuation of the current pace of asset purchases of $85bn per month on an open-ended basis, which would imply that the current $45bn per month in twist-financed Treasury purchases is replaced by $45bn per month in QE-financed Treasury purchases. It is possible that the committee will decide to lower the average maturity of the Treasury purchases slightly in order to avoid raising their ownership of specific longer maturities too much. They could do this without reducing the pace of duration removal, as they would be financing the purchases with base money creation (which has zero duration) instead of sales of shorter-dated Treasuries (which have positive duration).
Q: Will the committee move to an Evans-type rule, or some other type of outcome-based forward guidance?
A: In principle, there is widespread agreement that outcome-based guidance would be preferable, for the reasons laid out by Michael Woodford in his 2012 Jackson Hole paper and discussed at length in our own research. The most likely form of outcome-based guidance would be some version of the "Evans rule," defined as a commitment not to raise the federal funds rate until the unemployment rate had fallen to a particular threshold unless inflation had risen above a particular threshold. (Fed officials would undoubtedly be careful to specify that unemployment below the threshold would be a necessary but not sufficient condition for a rate hike; it is less clear whether inflation above the threshold would be characterized only as a necessary condition for a hike or also a sufficient one.)
However, the practical obstacles to such a move are still significant. The biggest issue is that the absolute level of the unemployment rate may not be a good indicator of the employment side of the mandate. This is partly because there is uncertainty about the structural unemployment rate, and partly because the "participation gap"--the difference between the actual labor force participation rate and the estimated structural participation rate--may embody almost as much labor market slack as the unemployment gap. This means that the unemployment rate alone does not pin down the overall amount of labor market slack very precisely. One could in principle address this by formulating the outcome-based guidance in terms of a combination of the unemployment rate, the participation rate, and perhaps other labor market measures. But in practice this would be very difficult to communicate and would risk introducing substantial volatility into the market's monetary policy expectations and thus broader financial conditions.
The implication is that we suspect the committee is stuck with the calendar-based guidance for some time yet. Nobody really likes it, but that has been true ever since the "mid-2013" language was adopted at the August 2011 meeting, and a better alternative may continue to prove elusive. We will probably learn more in the minutes of the October 23-24 meeting.
Q: Will the committee move to a consensus economic forecast?
A: The other innovation under discussion is the adoption of a consensus economic forecast by the committee as a whole. The committee has now conducted two experimental exercises intended to shed light on the feasibility of adopting a consensus economic and monetary policy forecast, and a broader discussion was on the schedule for this meeting. However, there are several questions surrounding such a move that have not yet been cleared up.
First, who would drive the consensus forecast? Would it be based on a proposal by the chairman developed in conjunction with the Board staff, or a more democratic process that involves initial input from all participants?
Second, what would be the monetary policy assumptions underlying such a forecast? Would it be based on constant interest rates, market expectations, or the committee's own expectations determined simultaneously with the economic forecast?
Third, how would FOMC participants indicate their personal agreement or disagreement with the consensus forecast? These are all knotty questions to answer, and we believe that it will be difficult to agree on moving to a consensus forecast anytime soon. However, the minutes will again provide more information on where this discussion stands.